I recognise the fast-growing hype very clearly when my Swedish colleague, Johanna Bernshed, and I are meeting enterprises all over the Nordics. (By the way, the photo above is of course the two of us, visiting a customer in Northern part of Sweden).

While the hype about Speech Analytics is increasing, still more companies ask us for help to get started. But I am also perfectly aware, that you could be somewhat confused by this new phenomenon. This article is meant as an introduction of Speech Analytics to you.

What is Speech Analytics: A short introduction

Speech Analytics is the process of analysing recorded calls to gather customer information to improve communication and future interaction. The process is primarily used by customer contact centres to extract information buried in interactions between customers and the enterprise.

This will be replacing decisions based on pure gut feelings and individual person’s more or less personal perceptions. Introducing data based insights of EVERYTHING that has to do with your business and customer interactions, resulting in a foundation of decision making based on data.

 This will be replacing decisions based on pure gut feelings and individual person’s more or less personal perceptions.

Speech Analytics is relevant to most functions and roles in an enterprise, employing from 30 agents and upwards. It is as relevant to the management, the business development, BI, Marketing, online-development as to the contact centre itself.

It’s all about automated transcription of your calls

Speech Analytics is basically a system that transcribes your calls. That is also why you must record the calls to make use of Speech Analytics. By transcribing your calls, you convert the audio files into text – and by doing that into data.

Depending on which Speech Analytics system you chose, you can use these transcriptions in various ways. The simplest systems gives you the option of searching calls, where certain specific terms are mentioned.

This can be very helpful, but not by any means as beneficial as the most advanced systems.

Advanced systems like Verint allows you to investigate your calls in almost endless ways providing you with both very operational and strategic insights. It is not an understatement, that having this kind of tool is quite addictive to anyone occupied with any kind of business improvement.

 The simplest systems gives you the option of searching calls, where certain specific terms are mentioned. This can be very helpful, but not by any means as beneficial as the most advanced systems.

What to focus on, while choosing a Speech Analytics system

There are some important differences between the different Speech Analytics systems, so here is 3 things to what to look out for (except of course for the ability to transcribe in a high quality):

  • Ability to categorise calls: As all human beings, we all tend to express ourselves in very different ways. Look for a Speech Analytics system, that helps you group all calls about the same issues, even if customers and agents use different wordings.
  • Ability to analyse the full transcription: Look for a Speech Analytics system that not only gives you what you are looking for, but also tells you about things, you didn’t know to look for. A system, that investigates the full transcription for patterns and trends and ‘google’ words, that are not yet categorised.
  • Part of a WFO suite: Some Speech Analytics vendors are also developing a series of related solutions like e.g. Text Analytics, Automated Quality Management and Performance Management – also known as a Workforce Optimisation (or WFO)- suite. Choosing a Speech Analytics system, that is a part of a WFO suite, just gives you far more options and ways to benefit from your Speech Analytics data.

10 reasons why Speech Analytics is hyped right now

All enterprises and organisations have their differences. They have different ways of working, different goals and different strategies. But after visiting a wide range of Nordic companies during the last years, it strikes me, that most of them also do have a range of profound pain points in common.

Pains, that Speech Analytics is helping to ease or reduce.

Here is my list 10 of the most frequent pain points, that I experience at our customers. I believe you can relate to at least a few of them:

1.     Black box of customer calls: What is really going on in your customer calls appears to be a big black box. You might listen to a sample of the calls already, but even this task is too massive and what exactly can you conclude by listening to only a sample? Why are the customers calling?

2.     Customer surveys: You are doing (collecting) customer satisfaction surveys, but only some customers are answering and consequently you don’t know, how they are experiencing your service.

3.     Net Promoter Score: Your NPS (Net Promoter Score) has ’frozen’ at a certain level, and you are missing ideas to make a progress. What to do?

4.     Call logging: You ask your agents to log the call-reasons, but it is time consuming and makes the average handling time increase – and the logging data does not seem to be trustworthy.

5.     The statistics you get from your call logging is superficial. Basically you don’t really know, why your customers are calling you.

6.     Agent feedback: When providing your agents with feedback, whether that is based on call listening or customer feedback surveys, you can’t always tell, if the feedback is fair to the agent.

7.     ChatBot: You are programming a new chatbot, but you miss information of what exactly the chatbot should communicate to be relevant to your customers (and bring down the number of calls).

8.     Campaigns: You basically don’t know the efficiency of your marketing or communication campaigns. Rumours are, that there is some reaction, but to what extent do your customers actually react? And how?

9.     Self-service: You don’t know if your contact centre agents are promoting self-service, asking for marketing permission or whatever you want them to do.

10.  First call resolution: You don’t know, to what extent your agents manage to do ‘first call resolution’ or how to improve this.

Try Speech Analytics on your own calls

There are different ways to demonstrate a Speech Analytics system. At Axcess Nordic, you will have the option of getting a full installation at your contact centre for a limited period. During this pilot test period we are letting you test the system on a wide number of your calls – it can even be live calls in production, we are testing.

It doesn’t matter if that number is 10.000, 100.000 or even 1 million calls. By doing this, you will get deep insight into a very large pool of unstructured data, and at the same time have a unique option of getting a true hands on experience with Verints Speech Analytics system. This will provide you with a perfect foundation to decide if Speech Analytics should be your source of insight going forward.

The insights, you will get during this pilot test, are yours to keep anyway.

If you want to know more, don’t hesitate to contact me or my Swedish colleague, Johanna Bernshed.

Read more about Speech Analytics in SwedishDanish or English at the Axcess Nordic homepage.

I am a Speech Analytics specialist at Axcess Nordic. Axcess Nordic helps service-oriented companies optimise and develop customer contact. I can be contacted via LinkedIn or email 

About us

We help service-oriented companies and organisations throughout the Nordic countries to optimise and develop their customer experience. At the same time, we help Nordic finance houses and energy companies with telephony and information systems.

Axcess Nordic A/S

Lautrupvang 6, DK-2750 Ballerup


+45 70 26 41 41

CVR: 34082863

Sales and delivery terms

Follow us:

Contact us